Shallow Water Depth Inversed Using Multispectral Satellite Based on Machine Learning

Jinlu Liu; Deyong Sun; Deyu Kong; Xishan Pan; Hongbo Jiao; Zhenghao Li; Shengqiang Wang; Yijun He

doi:10.3788/LOP220584

Laser & Optoelectronics Progress, Volume. 60, Issue 10, 1028007(2023)

Shallow Water Depth Inversed Using Multispectral Satellite Based on Machine Learning

Jinlu Liu¹, Deyong Sun^1,2、*, Deyu Kong³, Xishan Pan³, Hongbo Jiao⁴, Zhenghao Li¹, Shengqiang Wang^1,2, and Yijun He^1,2

Author Affiliations

¹School of Marine Sciences, Nanjing University of Information Science & Technology, Nanjing 210044, Jiangsu , China

²Key Laboratory of Space Ocean Remote Sensing and Application, Ministry of Natural Resources, Nanjing 210044, Jiangsu , China

³Jiangsu Provincial Marine Environment Monitoring Engineering Technology Research Center, Nanjing 210044, Jiangsu , China

⁴National Marine Data and Information Service, Tianjin 300171, China

show less

Abstract Get PDF(in Chinese)

Figures & Tables(13)

Fig. 1. Study area. (a) South Andaman islands; (b) Kume-jima; (c) Mentawai islands; (d) Hateruma-jima

Download full size

Fig. 2. Technology roadmap

Download full size

Fig. 3. Schematic of BP neural network

Download full size

Fig. 4. Schematic of random forest

Download full size

Fig. 5. Comparison of modeling accuracy. (a) MLR; (b) BP neural network; (c) RF

Download full size

Fig. 6. Comparison of validation accuracy. (a) MLR; (b) BP neural network; (c) RF

Download full size

Fig. 7. Residual distribution. (a) MLR; (b) BP neural network; (c) RF

Download full size

Fig. 8. Precision comparison between sub-region model and whole-region model

Download full size

Fig. 9. Inversion result. (a) Hateruma-jima; (b) Kume-jima; (c) South Andaman islands; (d) Mentawai islands

Download full size

Table 1. Data source introduction

View table

Table 1. Data source introduction

Sensor	Wavelength /μm	Central wavelength /μm	Resolution /m	Imaging time
Landsat-8 OLI	Coastal：0.430-0.450（B₁）	Coastal：0.440	30	Hateruma-jima 2013-06-05 Kume-jima 2015-12-13 South Andaman 2021-03-10 Mentawai 2019-05-02， 2019-05-27，2019-01-26
	B：0.450-0.510（B₂）	B：0.480
	G：0.550-0.590（B₃）	G：0.570
	R：0.640-0.670（B₄）	R：0.655
	NIR：0.850-0.880（B₅）	NIR：0.865
	SWIR1：1.570-1.650（B₆）	SWIR1：1.610
	SWIR2：2.110-2.290（B₇）	SWIR2：2.200

Table 2. Comparison of modeling accuracy of all models
View table
Table 2. Comparison of modeling accuracy of all models
Model MAE /m MAPE /% R²
MLR 2.37 35.48 0.68
BP 1.93 17.12 0.76
RF 1.01 9.02 0.93

Table 3. Comparison of validation accuracy of all models
View table
Table 3. Comparison of validation accuracy of all models
Model MAE /m MAPE /% R²
MLR 2.45 25.38 0.66
BP 2.05 19.43 0.74
RF 1.94 18.29 0.75

Table 4. Precision comparison of sub-region model and whole-region model

View table

Table 4. Precision comparison of sub-region model and whole-region model

Index	Region	South Andaman	Hateruma-jima	Kume-jima	Mentawai
R²	Sub-region	0.74	0.87	0.71	0.64
R²	Whole-region	0.74	0.87	0.72	0.66
MAE /m	Sub-region	1.95	1.34	2.30	1.78
MAE /m	Whole-region	1.95	1.36	2.11	1.80
MAPE /%	Sub-region	18.32	11.60	23.03	18.12
MAPE /%	Whole-region	18.37	11.40	21.19	16.79

Tools

Get Citation

Copy Citation Text

Jinlu Liu, Deyong Sun, Deyu Kong, Xishan Pan, Hongbo Jiao, Zhenghao Li, Shengqiang Wang, Yijun He. Shallow Water Depth Inversed Using Multispectral Satellite Based on Machine Learning[J]. Laser & Optoelectronics Progress, 2023, 60(10): 1028007

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites